Salud Pública

Age-related hearing loss is a potentially modifiable risk factor for cognitive impairment and dementia. We related a measure of the severity of hearing loss, the pure tone average (PTA) hearing threshold on audiometry, to measures of brain aging, vascular injury, cognitive function, and incident dementia outcomes.

Framingham Heart Study (FHS) participants (n = 1656 [mean age: 58 years]) attending their 6^th quadrennial examination, who were free of stroke and dementia, were evaluated for hearing loss (HL), and underwent subsequent brain MRI and cognitive assessment, were studied. PTA for hearing loss was calculated as the average of air conduction thresholds at four frequencies (0.5, 1, 2, 4 kHz). We defined HL as a PTA threshold > 25 dB in the better ear, and we additionally considered four PTA threshold categories (range, in dB HL): Normal [0,16), Slight [16,26), Mild [26-40), and Moderate or greater (>40) dB. We used multivariate linear regression and related PTA and HL to MRI and cognitive measures, to change in these measures over 4 - 8 years of follow-up and to incident all-cause dementia, using Cox proportional hazards regression. The covariates assessed were sex, age, education level, and APOE-4 status.

At least mild HL was associated with smaller brain volume (p = 0.02) by an amount equivalent to 1.75 years of aging at a baseline age of 60 years, and with a decline in executive function on follow-up (p = 0.009]). Over the same time period, higher PTA was associated with increased white matter hyperintensity volume (p = 0.048). Slight or greater HL (³16 dB) was associated with a higher risk (71%) of developing all-cause dementia (Hazard Ratio [HR]:1.71, 95% confidence interval [CI]: (1.01, 2.90), p = 0.04), over 15 years of follow-up. This association was driven by participants with at least one APOE-4 allele (HR:2.86, 95%CI: (1.12, 7.28), p = 0.03).

Midlife hearing loss was associated with smaller brain volumes, accelerated decline in executive function, accelerated accrual of white matter abnormalities, and a near doubling of dementia risk, suggesting a potential causal or biomarker role for midlife hearing loss in the prevention of late-life dementia.